Enhancing analysis of diadochokinetic speech using deep neural networks

Yael Segal-Feldman, Kasia Hitczenko, Matthew Goldrick, Adam Buchwald, Angela Roberts, Joseph Keshet

Abstract: Highlights•DDK tasks are often used to assess neurodegenerative disease using voice.•Manual analysis is time-consuming and offers a rough depiction of speech.•This paper introduces a state-of-the-art open-access model.•The model performs automatic segmentation of stop consonants and vowels.•The model utilizes the raw waveform data for analysis.